If you look at a site/product like Lumosity.com, you can compare your results to other age groups. If you accept that the tests on Lumosity do test particular brain functions/areas then its clear to see that from the millions of user that take part, the performance average does decline with age. It's actually quite dramatic to see that a 40 year old that scores in the 90% in a test category, fits in at 50% among the 20 year olds.
You can't really get any meaningful results from a self selected sample.
Since there is no other evidence that a 40 year old in the 90th percentile for mental performance is equal to a 20 year old in the 50th percentile, it seems more than likely that the data is skewed.
Perhaps Lumosity.com attracts above average young people and below average older people (given those results, some variation on that theme is the likely explanation).
An otherwise almost imperceptible decline in mental function could have have a huge impact on ranking among top chess players simply because at the top levels the margins between players are so small.
If chess Grandmasters dropped in mental acuity as fast as those Luminosity results suggested, a 60 year old Grandmaster could be beaten by any young novice.
Also worth mentioning is that competitive chess is favors very quick thinking and action in a way that most careers don't.
